Shifts of coral reefs towards alternative states occur due to local and global stressors. Although global stressors are expected to increase due to climate change, anthropogenic local stressors can be addressed to prevent the loss of important ecosystem services. Identifying and understanding how human activities affect the dynamics in the benthic communities in the reef ecosystems could facilitate more effective reef restoration efforts. But how do human activities affect water quality and subsequently the benthic cover? To answer this question we look towards the coral reefs of Bonaire, home to one of the most pristine reefs in the Caribbean. We combine existing data on human activities and environmental variables with new temporal water quality and benthic cover data generated along the west-coast of Bonaire. We created two sets of models: relating the benthic cover to water quality and explaining water quality with human activity. Because our data collection extended into unexplored deeper parts of the reef we have a unique opportunity to consider the effect of local stressors along a more extensive depth gradient.
We hypothesized that areas with high nutrient loads would be reflected by benthic cover with relatively high algae, sponges and benthic cyanobacterial mats. Our results showed this to be the case for sponges and turf algae, but not for benthic cyanobacterial mats (BCM) and macroalgae. The coral and crustose coralline algae (CCA) cover were expected to be related negatively to the selected water quality variables. The models predicting the coral cover give a mixed result. Both significant positive and negative effects of nutrients on the coral cover have been found, and relatively the positive effects are stronger than the negative effects. The expectation that an increase in human activity leads to a decrease in the water quality is a lot more nuanced, but it is clear that terrestrial human activity plays an important role. The influence of depth on the effects of the water quality on the benthic covers seems to be minimal at most. As the few significant differences in water quality effects found, were more likely to be the effect of under sampling than anything else. However, these results might change as more data becomes available, narrowing both the prediction and confidence intervals and thus increasing the chance of finding significant effects of water quality on the benthic cover and clearer effects of human activity.